The goals of this study are to determine how S. cerevisiae kinetochores bind to microtubules and how the forces required for chromosome movement are generated. This will be accomplished by linking systematic experiments to quantitative analysis. Stochastic models of yeast kinetochores and their ten known MAPs and motors will be generated using automated procedures and then calibrated to live-cell data using the method of indirect inference. Chromosome segregation is the process by which replicated DNA molecules are pulled into daughter cells by attaching to and moving along the microtubules of the mitotic spindle. Chromosome-microtubule attachment is mediated by kinetochores, multi-component protein complexes that form on centromeric DNA. S. cerevisiae is an attractive organism in which to study kinetochores because it contains the simplest of all known centromeres and because its molecular genetics is uniquely powerful. The conservation of kinetochore proteins from yeast to man suggests that principles learned from the study of simpler structures in yeast will be directly applicable to higher cells and to the genomic instability which aberrant kinetochore function causes. Aim 1: For kinetochore gene mutations, autoregressive moving average (ARMA) modeling and fuzzy clustering will quantify phenotypic similarity and heterogeneity. Aim 2: The role of tension in controlling kinetochore-microtubule interaction will be explored genetically and with ARMA having an input for exogenous variables (ARMAX) Aim 3: Stochastic models of kinetochores based on biophysical prior knowledge will be evaluated and calibrated to single-cell data using indirect inference and ARMA descriptors as a source of auxiliary parameters.We propose to study the cellular machinery responsible for dividing chromosomes, the DNA and protein structures that encode and preserve the genomel, into two equal sets upon cell division. Normally this process of "segregation" is very accurate, but it goes awry in cancer and gives rise to the genetic plasticity that makes long-term treatment of cancer problematic.